Longitudinal epidemiological studies on inequality and its impact on child development

Parental traits, behaviour and characteristics are among the best predictors of children's mental wellbeing. However, despite years of research and debate, there is no consensus on the causal interplay between parental characteristics and child outcomes. The question is highly relevant to define appropriate interventions to promote a healthy child development. The most studied parental characteristic is education. Parental education has been related to a broad range of health outcomes in the offspring, such as mortality, mental health, suicide attempts as well as social outcomes such as socioeconomic status. Nonetheless, evaluating whether and to what degree the association between parental characteristics and child outcomes is causal, is challenging. Child behaviours emerge through complex developmental processes influenced by biological, psychological, social, and environmental characteristics of the parents and characteristics of the child. Ethical constraints prevent us from developing experimental studies to evaluate child development when exposures present a risk for harm or disadvantage. Thus, we have to rely on observational data to draw causal inferences of such associations.
In this thesis I aimed to study the relation between characteristics of the parents and child development, in chapters 2, 3 and 4. I studied the relation between characteristics of the child and subsequent child development, in chapters 5, 6 and 7.
The studies presented in this thesis use data from three prospective population-based child cohort studies. Generation R Study is conducted in Rotterdam, the Netherlands. A total of 8,879 pregnant women, expecting delivery between April 2002 and January 2006, participated in the cohort. The Millennium Cohort Study (MCS) is a nationally representative study conducted across the United Kingdom. The cohort enrolled 18,552 families with 18,827 children between 2000 and 2001. Data collection started at age 9 months, data used in this thesis comprise the first sweep to the examination at child age 11 years. Growing up in Scotland (GUS) is a research initiative designed to monitor the experiences of three cohorts of children. The sample is nationally representative of Scottish population. The data used in this thesis comes from the birth cohort ensembled in 2005. This cohort enrolled 5,217 children born between June 2004 and May 2005. Data was collected annually from their families starting at age 10 months.
In Chapter 2, I explored the mechanisms that contribute to the correlation between parental education and children's academic success. Our emphasis was on parenting practices and child intelligence as potential mediators, recognizing that these aspects may vary slightly over time, thus needing repeated assessments, and taking into account that mediators can also be confounders. Child intelligence in early and mid-childhood mediated around 40% of the association of parental education with children’s school achievement, reflecting the role of genetic and environmental factors in this association. Parenting practices, namely family routines in early and mid-childhood, also mediated the relation of parental education to child school achievement. Routines provide predictability in the child’s environment fostering a stable emotional climate. These routines seem to support children to acquire academic competencies. We further evaluated the joint (dependent) mediated effect of parenting and intelligence in early and mid-childhood, but there was no evidence of a sequential mediation in which the mediators determine each other.
Chapter 3 evaluates the association of maternal educational mobility with parenting and offspring school achievement, intelligence, and genetic endowment. As maternal mobility cannot cause child genetic endowment, I suggest this association reflects genetic confounding. Upward maternal educational mobility was associated with better school achievement, a higher intelligence score in children and with more family routines. However, children of mothers with upward educational mobility did not reach the same school achievement, intelligence and family routines as children of mothers with a similar level of education but without mobility (so the mother’s education level was similar to grandparental education). This disadvantage of children of mothers with educational mobility suggests that assessing social inequalities as defined solely current parental socioeconomic status might not be sufficient to achieve equality. In other words, there are factors determining children's school achievement, children's intelligence and family routines that are (partially) independent of maternal education. For instance, parental practices are culturally inherited and likely transmitted across generations. We also found that maternal educational mobility was associated with child genetic endowment of academic attainment when compared with lower education and the same grandparental education. When maternal education mobility was compared to mothers with the same education but higher grandparental education, children of higher educated grandparents had a more favourable child genetic endowment of academic attainment. This hints at a role of genetics in educational mobility.
In chapter 4, I examined the relations of children’s home learning environment and parental mental health with language competence, unravelling the mediating role of social and emotional adjustment in these relations. I found that child mental health problems explained the association between maternal depressive symptoms and child structural language competence, whereas child mental health problems only partially mediated the association between maternal depressive symptoms and child pragmatic language skills. Child mental health problems also mediated the association between the home learning environment and child pragmatic language skills, while the symptoms only played a modest role in the association of the home learning environment with child structural language ability. This study is an example of how parental traits and the environment they provide are linked with developmental outcomes across childhood.
Chapter 5 describes bidirectional associations between internalizing symptoms, exter-nalizing symptoms and language ability. Our analysis supported the early onset, comorbidity and persistent nature of internalizing symptoms, externalizing symptoms and language ability. This may reflect the presence of common risk factors for the development of inter-nalizing and externalizing symptoms and language ability early in life. In addition, externalizing symptoms at ages 3 and 5 were negatively associated with language ability development. This finding suggest a compounding effect of mental health problems over time: not only do problems start at an early age and tend to be comorbid, children with mental health problems seem to fall behind in development. Better language ability at age 7 was related to less internalizing and externalizing symptoms at age 11. As such, language ability can provide the tools to manage emotions, build supportive relationships, solve problems, and achieve academic success.
In Chapter 6, I assessed the relation of childhood aggressive behaviour, emotional, and attention problems before school entry, with academic attainment at the end of primary education. I found that attention problems across childhood, but not aggression or inter-nalizing symptoms, were related to a poorer school achievement, which highlights the need to address attention problems at schools. Furthermore, internalizing symptoms before age 12 were related to a better school achievement at the end of primary education when accounting for the co-occurring externalizing symptoms. A detailed study of the mech-anisms of how internalizing symptoms relate to better school achievement could provide valuable insights into potential protective factors and coping strategies that enhance academic performance in children.
Chapter 7 investigated whether child mental health problems associate with academic under- and overachievement. I found that attention problems, independently of other mental health symptoms, related to a poorer school achievement than expected based on child intelligence. This finding is in line with findings on children with attention deficit hyperactivity disorder and highlights that attention problems even at a subclinical level can have negative implications for child development.
Finally, Chapter 8 presents a comprehensive discussion of the results, offering a complete interpretation of the studies, addressing key methodological considerations, and delineates the significance along with suggestions for future research.